Acknowledgements

Manda is built on the shoulders of open-source projects and data. The libraries and datasets listed below are used in the application and are redistributed or accessed at runtime.

Runtime libraries & data

CC-CEDICTCC BY-SA 3.0

CC-CEDICT contributors / MDBG

Chinese–English dictionary dataset that powers all word lookups.

Unihan Database (Unicode)Unicode License

Reading-frequency data used to pick the most likely pinyin for a polyphonic character: kHanyuPinlu corpus counts, with kMandarin for rarer characters. CC-CEDICT lists a character’s readings but not how common each one is.

Hanzi WriterMIT

Character stroke-order animation used on word detail pages.

Make Me a Hanzi / ArphicCC BY-SA 4.0 (Arphic Public License)

Arphic Technology Co., Ltd. / Make Me a Hanzi contributors

Stroke-path data used by Hanzi Writer, and the source of character component decomposition data (char-components.json) shown on Word Detail pages.

jieba-wasmMIT

WebAssembly port of the Jieba Chinese word segmenter, used to split text into words.

opencc-jsMIT / Apache 2.0

Converts between Simplified and Traditional Chinese character forms.

ReactMIT

UI library used to build the entire application.

React RouterMIT

Client-side routing between pages.

@dnd-kitMIT

Drag-and-drop toolkit used for reordering dashboard widgets.

Tesseract.jsApache 2.0

Pure JavaScript OCR engine running in a browser Web Worker; used to extract Chinese text from uploaded images.

foliate-jsMIT

Parses uploaded .epub files entirely in the browser (native DOMParser, no server), powering the Import Text page's Upload EPUB option (table of contents and chapter text extraction).

@zip.js/zip.jsBSD-3-Clause

Zero-dependency ZIP reader that unzips .epub files for foliate-js to parse.

HanziLookupJSMIT

Stroke-based Chinese character recognition used in the dictionary handwriting pad.

complete-hsk-vocabularyMIT

HSK 1–7-9 word lists (readings, definitions, classifiers) that power the HSK pages, Learning Path, and HSK badges.

TatoebaCC BY 2.0 FR

Bundled Simplified and Traditional Chinese example-sentence corpora (~48,000 / ~39,000 sentences), used on Word Detail pages, Cloze sessions, and Anki export.

@supabase/supabase-jsMIT

Client library for account sign-in, cloud sync, word comments, bug reports/feature suggestions, and moderation.

posthog-jsMIT / Apache 2.0

Product-analytics client used to capture anonymous usage events.

sql.jsMIT

WebAssembly build of SQLite, used to assemble the .anki2 database for Anki deck export.

fflateMIT

Zips the SQLite database and media manifest into a genuine .apkg file for Anki export.

ZodMIT

Schema validation used to safely parse attacker-controllable JSON on the Import/Export page before writing to storage.

Lingva TranslateAGPL-3.0

Privacy-respecting frontend for Google Translate, queried at runtime to auto-translate example sentences and reader text into English.

MyMemoryFree API

Translation memory API used as a silent fallback when Lingva Translate is unavailable.

Noto Sans/Serif SC, Noto Sans/Serif TC, Noto Sans HKSIL OFL 1.1

Simplified and Traditional Chinese typefaces, selectable as a font option throughout the app and served via Google Fonts.

ZCOOL XiaoWei & ZCOOL KuaiLeSIL OFL 1.1

Decorative Simplified Chinese typefaces, selectable as a font option, served via Google Fonts.

Ma Shan ZhengSIL OFL 1.1

Cursive brush-script typeface, selectable as a font option, served via Google Fonts.

LXGW WenKai TCSIL OFL 1.1

Traditional Chinese calligraphy-style typeface, selectable as a font option, served via Google Fonts.

Build tools

These tools are used during development and are not included in the distributed application.

ViteMIT

Build tool and development server.

Tailwind CSSMIT

Utility-first CSS framework used for all styling.

TypeScriptApache 2.0

Typed superset of JavaScript used throughout the codebase.